A New Physics-Based CO2 EOR Screening Tool for Offshore Applications

Abdulrahman Abdulwarith,Utkarsh Sinha, Sandarbh Gautam,Birol Dindoruk

Day 3 Wed, April 24, 2024(2024)

引用 0|浏览3
暂无评分
摘要
Abstract In this study we primarily focused on CO2 EOR tool development for offshore applications where well-distances tend to be larger than the onshore applications along with higher permeabilities. Furthermore, most of the offshore cases show a wider spectrum of variability in terms of fluid properties. Offshore reservoirs, especially the ones in deepwater, given their scale and high Original Oil in Place, are receiving significant attention for CO2 injection for EOR/Storage applications. Existing screening criteria, largely qualitative and/or averaged over the domain and fail to incorporate the key reservoir properties, fluid properties, reservoir heterogeneity and operational conditions. In this study, we introduce a new comprehensive physics-based CO2 screening tool for offshore applications, capturing first-order reservoir and fluid properties, heterogeneity (layering), and key operational parameters such as injection rates, pressures and well distances. We used a wide spectrum of reservoir fluid data, as the fluid properties are one of the dominant control parameters for CO2 injection. The fluid information provided was limited to C7+ and they are clustered into six representative subgroups using the K-means algorithm. Despite limited information on the fluid data, especially in the context of calibration for CO2 floods, we conditioned outcomes using only C7+ composition, utilizing estimators like Machine Learning Based Minimum Miscibility Pressure from our previous study (Sinha et al. 2021). A set of base cases for reservoir simulation were defined, and several simulations runs were performed considering reservoir heterogeneity/layering using different Dykstra-Parson's coefficients, dip angle and operational parameters such as CO2 injection rates and the well distances. Using systematic and exhaustive set of simulation runs, we developed a predictive model based on K-nearest neighbor algorithm to predict the performance of CO2 injection in terms of incremental production, recovery factor, and breakthrough time and the base case runs were also compared against the outcome of the CO2 injection per case defined using the tool/methodology developed. The screening tool developed was examined using various combinations of in-situ fluid compositions, different degrees of heterogeneities defined by Dykstra-Parson's coefficients (DP), dip angles, CO2 injection rates, and well distances. The results showed a high level of agreement between the tool's predictions and the outcomes from the original reservoir simulation runs and variants. Furthermore, the screening tool provided insights into the performance of CO2 injection at selected times by the user, focusing on metrics like pore volume injected versus the recovery factor, cumulative oil production, and the percentage of CO2 at the production well. This dynamic (time-dependent) estimation of CO2 injection performance enables more effective flood evaluation, reservoir surveillance and management, allowing for the selection or modification of optimal operational parameters to maximize the incremental oil produced and delaying the CO2 breakthrough. Based on our sensitivity analysis concerning the injection rate, we found that high injection rates of 20 MMSCF/D led to earlier CO2 breakthrough and a reduced recovery factor compared to lower injection rates of 5 MMSCF/D. Additionally, in terms of the heterogeneity effect (layering defined by DP coefficient), a higher degree of heterogeneity (indicated by high DP coefficients) resulted in lower recovery, as anticipated. However, for heavier fluids, increased layering led to a better recovery factor due to vertical cross flow due to the geometries considered. CO2 screening tool developed for offshore CO2 injection applications is the first in this area to go beyond simple screening methodologies and is easily deployable while lending itself for further development for various input data set combinations. This tool pioneers the incorporation of the dominant factors capturing the physics of the flow reservoirs and provides a quantitative assessment of CO2 injection performance. Hence, the developed tool represents a significant advancement for CO2/EOR applications in offshore settings while can be adjusted to other reservoir conditions/settings.
更多
查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要